PDF )   Rev Osteoporos Metab Miner. 2021; 13 (1): 10-16
DOI: 10.4321/S1889-836X2021000100003

Gómez Alonso C1, Rodríguez García M2, Avello Llano N3, García Gil-Albert C3, Palomo Antequera C4, Fernández Villabrille S1, Rodríguez Carrio J5, Díaz Sottolano AA6, Fernández Martín JL1, Cannata Andía JB1, Naves Díaz M1
1 Bone Metabolism Clinical Management Unit. Central University Hospital of Asturias. University of Oviedo, Institute of Health Research of the Principality of Asturias (ISPA). Renal Research Network of the Carlos III Health Institute (REDinREN of the ISCIII). Oviedo (Spain)
2 Nephrology Clinical Management Area. Central University Hospital of Asturias. University of Oviedo, ISPA. REDinREN of the ISCIII. Oviedo (Spain)
3 Laboratory of Medicine. Central University Hospital of Asturias. Oviedo (Spain)
4 Unit of Clinical Management of Internal Medicine. Central University Hospital of Asturias. University of Oviedo, ISPA. Oviedo (Spain)
5 Basic and Translational Research in Chronic Inflammatory Diseases. University of Oviedo, ISPA. Oviedo (Spain)
6 Hospital de Cabueñes, ISPA. Gijón (Spain)

 

Summary
Objetive: Biochemical parameters continue to be the most widely used option for the follow-up of patients with bone metabolic disorders. The objective of our study was to assess the association of some biochemical markers of bone metabolism with the appearance and progression of aortic calcifications.
Material and methods: In this study, 624 men and women older than 50 years were selected at random. The participants completed a questionnaire and underwent two lateral dorsal-lumbar x-rays and bone densitometry. Four years later, the same studies were repeated in 402 subjects along with a biochemical study.
Results: Age and the proportion of men were higher in those who had “global progression” of aortic calcification (progression of the existing ones plus new ones). The serum levels of calcium and calcitriol were significantly higher and those of osteocalcin significantly lower in which “global progression” of aortic calcification was observed. Multivariate analysis showed that only osteocalcin was independently associated with “global progression” of aortic calcification, with an 18% decrease for each 1 ng/mL increase in osteocalcin levels (odds ratio (OR)=0, 82; 95% confidence interval (95% CI): 0.71-0.92). The categorization of osteocalcin into tertiles showed that the subjects of the first tertile (<4.84 ng/mL) were associated with a higher proportion of new aortic calcifications: (OR=2.45; 95% CI: 1.03-3, 56) with respect to the third tertile (>6.40 ng/mL).
Conclusion: Serum levels of osteocalcin could be a biochemical marker to evaluate the appearance and/or evolution of aortic calcification. However, it is necessary to determine with greater precision how it could exert this protective effect in the process of vascular calcification.

Key words: osteocalcin, vascular calcification, biochemical markers, bone mineral density.

 

Introduction
Atherosclerosis, arteriosclerosis, vascular calcification and osteoporosis are common age-related disorders associated with high morbidity and mortality [1,2]. Due to the increased life expectancy in the Spanish population, these disorders are expected to become more and more frequent in the coming decades. Although recent work has been carried out on the development of non-invasive techniques for the early detection of vascular calcifications, such as pulse wave velocity and non-contrast carotid ultrasound, serum biochemical parameters continue to be the most widely used option for monitoring patients with bone metabolic disorders [3-5].
Having easily accessible non-invasive tools such as biochemical markers allow for the adoption of therapeutic measures in order to mitigate the deleterious effect of bone loss. Taking into account that osteoporosis and vascular calcification share etiopathogenic mechanisms [6,7], some biochemical parameters used to study bone metabolism could serve as possible markers of vascular calcification. Therefore, this study aims to assess the association of some biochemical markers of bone metabolism with the appearance and progression of aortic calcifications.

Material and methods
This study used data from a European project designed to determine the prevalence of vertebral fracture (European Vertebral Osteoporosis Study – EVOS) [8], in which the Bone and Mineral Metabolism Service of the Central University Hospital of Asturias took part.
For this work, 308 men and 316 women over 50 years of age were selected at random from the municipal registry of Oviedo. Our protocol involved patients’ filling out a questionnaire on risk factors related to osteoporosis, two lateral dorsal-lumbar x-rays and a bone densitometry (DXA) (the radiographic study was not completed in only 2 cases), and collecting anthropometric measurements such as height and weight to determine body mass index (BMI). All subjects had sufficient stamina to go up two floors without an elevator and 99% lived in their own home.
After four years, they were invited to repeat the radiological study, bone densitometry, anthropometric measurements and respond to a questionnaire on risk factors for osteoporosis and a biochemical study. In the second control, 402 subjects participated (213 women and 189 men), of whom 335 agreed to carry out the biochemical study. A total of 67 subjects (16.7%) were excluded from the analysis because they had been treated for osteoporosis or their renal function was impaired with serum creatinine greater than 0.8 mg/dL in women and 1.1 mg/dL in men, respectively. All data were available at baseline and at 4 years in 262 subjects.

Evaluation of the progression of vascular calcification
Abdominal aortic calcification was evaluated by two independent investigators, and was defined and classified as grade 0 (absent), grade 1 (mild-moderate), and grade 2 (severe). Isolated punctate calcifications, a visible linear calcification in less than 2 vertebral bodies, or a dense calcified plaque were defined as mild-moderate calcification [9]. The presence of a visible linear calcification along at least two vertebral bodies and/or the presence of two or more calcified dense plaques was defined as severe calcification. The degree of intra- and inter-observer concordance in the analysis of the x-rays was 92% and 90%, respectively, with a Kappa coefficient of 0.78 and 0.73, data that indicate good reproducibility [9].
The progression of aortic calcification was determined by comparing the x-rays taken at baseline with those at 4 years. It was defined as “global progression” of aortic calcification when an increase in the magnitude of baseline aortic calcification coexisted with the appearance of new calcifications, comparing the x-rays at the outset with those done 4 years later.

Densitometric evaluation
Bone mineral density (BMD) was measured with a Hologic® QDR-1000 DXA densitometer (Hologic Inc., Waltham, Massachusetts, USA). In all cases, the anteroposterior lumbar spine (L2-L4) and the proximal extremity of the right femur were analyzed. For the evaluation of lumbar BMD, 4 subjects with marked degenerative osteoarthritis were excluded. The coefficients of variation (CV) were 1.2% and 1.9%, respectively9. The precision and quality control was performed daily with a lumbar spine phantom, with which a CV of 0.0±0.1% was obtained. In the fourth year, BMD was determined in the same areas used in the first study, and the percentage of change between both measurements was used to evaluate changes in BMD.

Biochemical analysis
In the baseline study, no biochemical study was carried out. At 4 years, a fasting blood and urine sample was taken from each subject participating in the study. Once the serum was separated, the latter and the urine were kept frozen at -80ºC until quantification. Serum calcium, creatinine, phosphorus, total alkaline phosphatase, and acid resistant tartrate phosphatase were measured using an autoanalyzer (Hitachi Mod. 717, Ratigen, Germany). The serum levels of calcidiol (25OHD) were determined by prior extraction with acetonitrile (IDS, Ltd., Bolton, United Kingdom), whose intra- and inter-assay coefficients of variation (CV) were 5.2% and 8.2%, respectively.
Levels of 1,25-dihydroxyvitamin D were measured by radioimmunoassay (IDS, Ltd.); intra- and interassay CVs were 6.5% and 9%, respectively. The intact PTH and total osteocalcin levels were measured by radioimmunoassay (Nichols Institute, San Juan Capistrano, California, USA). Intra- and inter-assay CV values were 2.6% and 5.8% for PTH and 4.5% and 5.1% for osteocalcin, respectively.
All the tests carried out followed the principles set forth in the Declaration of Helsinki and were formally approved by the Committee for Clinical Trials of the Principality of Asturias.

Statistic analysis
Data analysis was carried out using SPSS version 17.0 for Windows. The quantitative variables were analyzed by Student’s t test and the qualitative variables by chi-square.
Multivariate analysis was performed using logistic regression adjusting for age, sex and BMI, in those serum or urinary markers in which the univariate analysis was significantly associated with progression and/or appearance of new abdominal aortic calcification.
Pearson correlations were performed between those biochemical parameters that, at the multivariate level, showed a significant association with the percentage of change in BMD between both cross-sectional studies.

Results
The mean age of those who had “global progression” of aortic calcification (progression of existing vascular calcifications plus new vascular calcifications) was higher than the age of those in whom this situation was not observed (Tables 1 and 2). However, there were no age differences in those who only presented a new aortic calcification as a change in the control at 4 years (Table 3). The BMI was similar in those with aortic calcification, both in those in which the calcification progressed, as in those with new calcifications, or considering both variations together (Tables 1-3).

Male sex was significantly more frequent in those in whom progression of existing aortic calcifications and/or new aortic calcifications was observed. In contrast, there were no differences in the smoking habit (Tables 1-3).
Regarding the biochemical markers of bone metabolism, serum levels of calcium and calcitriol were significantly higher and those of osteocalcin significantly lower in those subjects in whom “global progression” of aortic calcification was observed (new calcifications plus progression of vascular calcification ) (Table 1).
The logistic regression analysis adjusted for age, sex and BMI showed that the only biochemical marker that was independently associated with “global progression” of aortic calcification was osteocalcin, showing that increases of 1 ng/mL were associated with a decrease in 18% in the progression of aortic calcification (odds ratio (OR)=0.82; 95% confidence interval (95% CI): 0.71-0.95) (Table 4). Age and male sex were also significantly associated with progression of vascular calcification (OR=1.05; 95% CI: 1.01-1.08 and OR=2.06; 95% CI: 1.20-3.54, respectively) (Table 4).

In the univariate analysis, serum osteocalcin levels were significantly lower and calcium levels significantly higher in those subjects in whom aortic calcification had progressed (Table 2). The logistic regression analysis adjusted for age, sex and BMI confirmed that osteocalcin was the only parameter that showed a significant association: increases of 1 ng/mL were associated with a 16% increase in the progression of aortic calcifications (OR=0.84; 95% CI: 0.70-0.99) (Table 4). Sex (OR=1.95; 95% CI: 1.01-3.76) and also age (OR=1.06; 95% CI: 1.02-1.10) were associated in this multivariate model (Table 4).
When only those subjects who presented a new aortic calcification were analyzed, serum levels of osteocalcin were found to be significantly lower and those of calcitriol, significantly higher (Table 3). The logistic regression analysis adjusted for age, sex and BMI confirmed that only osteocalcin showed a significant association: increases of 1 ng/mL were associated with a 20% appearance of new aortic calcifications (OR=0.80; 95% CI: 0.67-0.97) (Table 4). Male sex (OR=2.30; 95% CI: 1.15-4.59), but not age, was associated in this multivariate model (Table 4).
The categorization of serum osteocalcin levels into tertiles showed that the lowest tertile (osteocalcin <4.84 ng/mL) was the one that showed the highest proportion of new aortic calcifications (22; 42.3%). The second tertile (osteocalcin between 4.84 and 6.40 ng/mL) showed the same trend, but in a lower proportion (18; 34.6%), while the third tertile (osteocalcin >6.4 ng/mL) ) showed the lowest proportion (12; 23.1%). A logistic regression analysis adjusted for age, sex and BMI showed that the subjects of the first tertile were associated with a higher proportion (2.45 times) of new aortic calcifications: (OR=2.45; 95% CI: 1.03-3.56). There were no differences with those of the second tertile (OR=1.48; 95% CI: 0.611-3.56).
The bi-variate correlations between the percentage of change in BMD at the lumbar and femoral neck level and the serum levels of osteocalcin showed a negative and significant correlation. Higher values of osteocalcin were associated with a lower loss of BMD, while lower values of osteocalcin were associated with greater losses of bone mass (Figure 1A and 1B).

Discussion
Our results confirm that, of the biochemical markers analyzed, osteocalcin was the only marker associated with the appearance and progression of aortic calcifications independently of age, sex and BMI. A 1 ng/mL increase in osteocalcin decreased the “global progression” of aortic calcification by 18%, a protection equivalent to being 3-4 years younger.
Osteocalcin, a vitamin K-dependent protein, is the most abundant non-collagen component in the mineralized matrix of bone. It is not only produced by bone, but also by vascular smooth muscle cells that show a phenotype similar to osteoblasts [10]. It inhibits the precipitation of calcium phosphate and shows a strong affinity for hydroxyapatite [11]. Initially, it was thought that osteocalcin inhibited the growth of hydroxyapatite crystals [12] and limited bone formation [13].
Experimental studies have shown that decarboxylated osteocalcin can up-regulate nitric oxide synthesis in human endothelial cells with a protective effect against endothelial dysfunction. These findings support the opinion that decarboxylated osteocalcin is the biologically active form of the protein, with a protective function on the vasculature independent of its metabolic role, although more studies are required to confirm this fact [14].
Osteocalcin has been detected to a greater degree in calcified plaques and aortic valves than in healthy non-calcified vessels [15,16]. The level of osteocalcin mRNA reportedly increases between 8 and 14 times in calcified aortic plaques compared to healthy aortas [17]. The increase in total osteocalcin may occur as a result from the development of an osteogenic phenotype in atherosclerotic plaques [18]. However, this requires further validation. Recently, osteocalcin has been found to play a crucial role in arterial calcification mediated by Wnt/β-catenin signaling through increased oxidative phosphorylation, and this finding may have clinical implications [19].
However, studies in patients are inconclusive. A relatively recent meta-analysis included 46 studies that examined the association between osteocalcin and atherosclerosis [20]. Of the studies that analyzed the association between osteocalcin and carotid intima-media thickness (CIMT), four reported that higher levels of osteocalcin were associated with greater CIMT, four reported that higher levels of osteocalcin were associated with a lower CIMT, and three did not find any correlation. However, studies that examined mononuclear cells positive for osteocalcin or histological staining for osteocalcin showed that higher levels of osteocalcin were associated with an increase in markers of atherosclerosis and calcification [20]. Thus, it is suggested that osteocalcin could be a marker of the calcification process.
Our results show that, in the 4-year period between both cross-sections, both the presence of new aortic calcifications and their progression were associated with lower levels of osteocalcin, regardless of age, sex and BMI. It is noteworthy that the lowest tertile of osteocalcin (<4.84 ng/mL) was associated with a significant increase in new aortic calcifications: 2.45 (1.03-3.56) compared to subjects with serum osteocalcin levels higher than 6.4 ng/mL.
Kim et al. found similar results in Asian women to those in our study, with an inverse correlation between osteocalcin and vascular calcification measured by the Agatston score, even after adjusting for age [21]. Similar results have also been shown in other cross-sectional [22,23] and longitudinal studies, such as ours, in which raised levels of osteocalcin are found to be associated with less progression of abdominal aortic calcification [24]. These authors suggest that osteocalcin could be involved in the aortic calcification process indirectly by its action on insulin and insulin resistance. Fusaro et al. have recently observed in a population on dialysis that those diabetic patients with a higher prevalence of vascular calcification had lower serum levels of total and decarboxylated osteocalcin [25]. In fact, in a secondary analysis of our study, analyzing osteocalcin levels in those subjects diagnosed with diabetes, it was observed that the presence of diabetes (n=23) was associated with significantly lower levels of osteocalcin than those without diabetes (n=241) (4.89±1.80 ng/mL compared to 5.96±2.14 ng/mL; p=0.020).
It could also be conjectured that low levels of osteocalcin are associated with vascular calcification (VC) due to less bone remodeling, which could be a VC risk factor [26,27]. However, this possibility would not be supported by the results of this study, since the subjects with lower levels of osteocalcin and higher VC were those with lower BMD, which would be more indicative of high remodeling than low remodeling [28].
On the other hand, the usefulness of osteocalcin as a serum marker remains controversial. There is still a long way to go to define whether osteocalcin can be used as a diagnostic or detection tool in the appearance of VC. It is noteworthy that no study has differentiated between forms of osteocalcin when it comes to VC. Consequently, it is necessary to study the effect that carboxylated and decarboxylated osteocalcin could have in this environment, as well as to consider the mechanisms associated with the increase of osteocalcin in calcified tissue [5].
This study presents several limitations. First, osteocalcin determination was only carried out in the second cross section, which limits the associations found. Second, intact or total osteocalcin was determined without differentiating between carboxylated or decarboxylated. On the other hand, the evaluation of vascular calcification was carried out by simple X-ray imaging and not by more sensitive techniques. It is also possible that some of the people who attended the second check-up after 4 years would have done so because they were in a worse physical condition compared to those who did not attend it, although no clear selection biases were found29.
Despite these limitations, the study also has important strengths, such as the adequate response of the subjects who participated in the study, both at baseline (50%) [30] and at 4 years of the follow-up period (70%). The degree of reliability among observers for the assessment of vascular calcification supports its use as a diagnostic criterion. Finally, unlike other studies, this study was prospective, and not cross-sectional like most of those cited. This reinforces the validity of the results found and their greater degree of association.
Thus, although new studies are needed to confirm these results, this study seems to indicate that serum levels of osteocalcin could be a promising biochemical marker associated with the appearance and/or development of aortic calcification.

Acknowledgments: This research study has been partially funded by the European Study on Vertebral Osteoporosis (EVOS), European Union (1991–1993); European Prospective Osteoporosis Study (EPOS), European Union (BIOMED 93–95), BMHI-CT 092-0182 (1993–1997); Health Research Fund (FIS 94/1901-E); Retic REDinREN of ISCIII (RD06/0016/1013, RD12/0021/0023 and RD16/0009/0017); National R + D + I Plan 2008-2011, State R + D + I Plan 2013-2016, European Regional Development Fund (ERDF), Science, Technology and Innovation Plan 2013-2017 and 2018-2022 of the Principality of Asturias (GRUPIN14-028, IDI-2018-000152), Renal Foundation Íñigo Álvarez de Toledo (FRIAT). Augusto Antonio Díaz Sottolano has been funded by the Principality of Asturias Health Research Institute (ISPA). Sara Fernández Villabrille has been financed by IDI-2018-000152 and project FIS 17/00715 and Javier Rodríguez Carrio by a contract along with Juan de la Cierva and Sara Borrell.

Conflict of interests: The authors declare no conflict of interest.

 

Bibliography

1. Wilson PW, Kauppila LI, O’Donnell CJ, Kiel DP, Hannan M, Polak JM, et al. Abdominal aortic calcific deposits are an important predictor of vascular morbidity and mortality. Circulation. 2001; 103(11):1529-34.
2. Cummings SR, Melton LJ. Epidemiology and outcomes of osteoporotic fractures. Lancet. 2002;359(9319):1761-7.
3. Torii S, Arima H, Ohkubo T, Fujiyoshi A, Kadota A, Takashima N, et al. Association between pulse wave velocity and coronary artery calcification in Japanese men. J Atheroscler Thromb. 2015; 22(12):1266-77.
4. Zhang H, Du J, Wang H, Wang H, Jiang J, Zhao J, et al. Comparison of diagnostic values of ultrasound micro-flow imaging and contrast-enhanced ultrasound for neovascularization in carotid plaques. Exp Ther Med. 2017;14(1):680-8.
5. Arroyo D, Betriu A, Martinez-Alonso M, Vidal T, Valdivielso JM, Fernández E. Observational multicenter study to evaluate the prevalence and prognosis of subclinical atheromatosis in a Spanish chronic kidney disease cohort: baseline data from the NEFRONA study. BMC Nephrol. 2014;15:168.
6. Kiel DP, Kauppila LI, Cupples LA, Hannan MT, O’Donnell CJ, Wilson PW. Bone loss and the progression of abdominal aortic calcification over a 25 year period: the Framingham Heart Study. Calcif Tissue Int. 2001;68(5):271-6.
7. Schulz E, Arfai K, Liu X, Sayre J, Gilsanz V. Aortic calcification and the risk of osteoporosis and fractures. J Clin Endocrinol Metab. 2004;89(9):4246-53.
8. O’Neill TW, Felsenberg D, Varlow J, Cooper C, Kanis JA, Silman AJ. The prevalence of vertebral deformity in european men and women: the European Vertebral Osteoporosis Study. J Bone Miner Res. 1996;11(7):1010-8.
9. Naves M, Rodriguez-Garcia M, Diaz-Lopez JB, Gomez-Alonso C, Cannata-Andia JB. Progression of vascular calcifications is associated with greater bone loss and increased bone fractures. Osteoporos Int. 2008;19(8):1161-6.
10. Evrard S, Delanaye P, Kamel S, Cristol JP, Cavalier E. Vascular calcification: from pathophysiology to biomarkers. Clin Chim Acta. 2015;438:401-14.
11. van de Loo PG, Soute BA, van Haarlem LJ, Vermeer C. The effect of Gla-containing proteins on the precipitation of insoluble salts. Biochem Biophys Res Commun. 1987;142(1):113-9.
12. Hunter GK, Hauschka PV, Poole AR, Rosenberg LC, Goldberg HA. Nucleation and inhibition of hydroxyapatite formation by mineralized tissue proteins. Biochem J. 1996;317(Pt 1):59-64.
13. Ducy P, Desbois C, Boyce B, Pinero G, Story B, Dunstan C, et al. Increased bone formation in osteocalcin-deficient mice. Nature. 1996;382(6590):448-52.
14. Tacey A, Qaradakhi T, Brennan-Speranza T, Hayes A, Zulli A, Levinger I. Potential role for osteocalcin in the development of atherosclerosis and blood vessel disease. Nutrients. 2018; 10(10):1426.
15. Levy RJ, Gundberg C, Scheinman R. The identification of the vitamin K-dependent bone protein osteocalcin as one of the gamma-carboxyglutamic acid containing proteins present in calcified atherosclerotic plaque and mineralized heart valves. Atherosclerosis. 1983;46(1):49-56.
16. Levy RJ, Zenker JA, Lian JB. Vitamin K-dependent calcium binding proteins in aortic valve calcification. J Clin Invest. 1980;65(2):563-6.
17. Fleet JC, Hock JM. Identification of osteocalcin mRNA in nonosteoid tissue of rats and humans by reverse transcription-polymerase chain reaction. J Bone Miner Res. 1994;9(10):1565-73.
18. Johnson RC, Leopold JA, Loscalzo J. Vascular calcification: pathobiological mechanisms and clinical implications. Circ Res. 2006;99(10):1044-59.
19. Rashdan NA, Sim AM, Cui L, Phadwal K, Roberts FL, Carter R, et al. Osteocalcin Regulates Arterial Calcification Via Altered Wnt Signaling and Glucose Metabolism. J Bone Miner Res. 2020;35 (2):357-67.
20. Millar SA, Patel H, Anderson SI, England TJ, O’Sullivan SE. Osteocalcin, Vascular calcification, and atherosclerosis: a systematic review and meta-analysis. Front Endocrinol (Lausanne). 2017;8:183.
21. Kim KJ, Kim KM, Park KH, Choi HS, Rhee Y, Lee YH, et al. Aortic calcification and bone metabolism: the relationship between aortic calcification, BMD, vertebral fracture, 25-hydroxyvitamin D, and osteocalcin. Calcif Tissue Int. 2012;91(6):370-8.
22. Ogawa-Furuya N, Yamaguchi T, Yamamoto M, Kanazawa I, Sugimoto T. Serum osteocalcin levels are inversely associated with abdominal aortic calcification in men with type 2 diabetes mellitus. Osteoporos Int. 2013;24(8):2223-30.
23. Kanazawa I, Yamaguchi T, Yamamoto M, Yamauchi M, Kurioka S, Yano S, et al. Serum osteocalcin level is associated with glucose metabolism and atherosclerosis parameters in type 2 diabetes mellitus. J Clin Endocrinol Metab. 2009;94(1):45-9.
24. Confavreux CB, Szulc P, Casey R, Boutroy S, Varennes A, Vilayphiou N, et al. Higher serum osteocalcin is associated with lower abdominal aortic calcification progression and longer 10-year survival in elderly men of the MINOS cohort. J Clin Endocrinol Metab. 2013;98(3):1084-92.
25. Fusaro M, Gallieni M, Aghi A, Rizzo MA, Iervasi G, Nickolas TL, et al. Osteocalcin (bone GLA protein) levels, vascular calcifications, vertebral fractures and mortality in hemodialysis patients with diabetes mellitus. J Nephrol. 2019; 32(4):635-43.
26. Frazão JM, Martins P. Adynamic bone disease: clinical and therapeutic implications. Curr Opin Nephrol Hypertens. 2009;18(4):303-7.
27. Bover J, Ureña P, Brandenburg V, Goldsmith D, Ruiz C, DaSilva I, et al. Adynamic bone disease: from bone to vessels in chronic kidney disease. Semin Nephrol. 2014;34(6):626-40.
28. Ueda M, Inaba M, Okuno S, Maeno Y, Ishimura E, Yamakawa T, et al. Serum BAP as the clinically useful marker for predicting BMD reduction in diabetic hemodialysis patients with low PTH. Life Sci. 2005;77(10):1130-9.
29. O’Neill TW, Marsden D, Silman AJ. Differences in the characteristics of responders and non-responders in a prevalence survey of vertebral osteoporosis. European Vertebral Osteoporosis Study Group. Osteoporos Int. 1995;5(5):327-34.
30. Naves M, Díaz López JB, Virgós MJ, O´ Neill TW, Gómez C, Zaplana J, et al. Indices de participación y aspectos metodológicos de interés en un estudio de prevalencia de fractura vertebral en Asturias. REEMO. 1993;2(5): 29-32.